October 23, 2009

The Twitter Question

Our report, Twitter and Status Updating, Fall 2009, has garnered a lot of attention, some of it focused on the question we use to measure this activity:

“Do you ever use Twitter or another service to share updates about yourself or to see updates about others?”

In the spirit of openness that characterizes our approach to research, we thought we’d share our answers to the questions asked over the last day or two. We also want to share some of our ideas for how to improve the measurement of status updates as an activity.

What do you mean by “another” service? We left it intentionally vague so that the respondent could answer yes if they use Yammer, or even if they change their Gmail status. We want to capture an activity, not just a brand. Since the first days of the Project our primary goal has been to measure general activity online, rather than particular services or applications.

Occasionally, we do ask about particular companies’ products – which social network sites people use; which blogging service they use; which news sites they visit – because we think that helps flesh out a story we are trying to tell. And we sometimes add prompts to our question language, as we did with the “Twitter or another service” question in an attempt to make clearer to respondents what we are asking about.

Still, there are any number of commercial marketing firms that cover the horse race between companies about the size of their user population. We see our mission as broader than that, so we frame our normal questions about online activities in general, rather than specific terms. At the same time, we phrased this question in a way that acknowledged that Twitter is the dominant service, so we feel comfortable reporting the data the way we did.

Facebook originated status updates, so doesn’t that hopelessly muddle the picture? Yes and no. Facebook is the original status update site and indeed, they get 45 million status updates a day from their users.

The Twitter question appears in the middle of a series about various internet activities, a version of which we include in every tracking survey we put in the field. We always start the series by asking about email, but then have a menu of about 70 activities to choose from.  This survey included the following activities, which were asked in random order after the email question:

Use an online dating site

Listen to music online at a website for a radio station, music store, recording artist or music service

Research your family’s history or genealogy online

Create or work on your own online journal or blog

Use a social networking site like MySpace, Facebook or LinkedIn.com

Take material you find online – like songs, text or images – and remix it into your own artistic creation

Share something online that you created yourself, such as your own artwork, photos, stories or videos

Use Twitter or another service to share updates about yourself or to see updates about others

Visit virtual worlds such as Second Life

Create or work on your own webpage

Create or work on web pages or blogs for others, including friends, groups you belong to, or for work

Post comments to an online news group, website, blog or photo site

Possible tweaks: We could pair up the social network site question with the Twitter question so the difference is clear to the respondent. We could also ask social network site users specifically about whether they post status updates, which we don’t do in this general activity battery.

If you know the question isn’t perfect, why not wait until it is? One of our biggest challenges over the last 10 years has been to capture an activity in a survey question as it is on the rise but before it morphs into something new. It is especially difficult when a platform is changing along with user practices around a certain activity. We have a good track record of getting questions about new activities into the field, testing them, tweaking them, and getting fresh data out into the public conversation and in this case we tested the wording of this question in three surveys before we released any results.

Can this question be saved? If we stick with the current wording, we are considering adding follow-up questions:

1) For people who answer “yes,” ask a multiple response follow-up about whether they post on Twitter, a social networking site, or someplace else. The “someplace else” could be left as a generic Other, an open end response, or a list of other services if we can come up with a decent
list.

2) For people who post updates in multiple places, it would also be interesting to know whether they post individually in each place, or whether they have it set up so that one post goes to multiple sites – or vice versa, where they post in multiple sites and have all their updates fed into one aggregator service like Friendfeed. This could also give us new insights into identity management.

3) Asking a third follow up for all those who say yes that asks whether they expect someone to respond to their posting or not, to better understand the expectations of the respondents about interaction in the various sites of status updating online.

We are also continually adding to and tweaking our battery of questions related to mobile internet use.  We could add a status update question to that battery to explicitly ask wireless internet users about this activity in the context of other mobile questions.

Keep it up – we love the feedback!  Please email us, contact us on Twitter (either @SusannahFox or @Pew_Internet), or give us a call: 202 419 4500. We want to know what you think.